Mercurial > libervia-backend
annotate src/plugins/plugin_xep_0071.py @ 771:bfabeedbf32e
core: i18n refactoring:
- _() is no more installed in __builtin__
- instead, there is a new sat.core.i18n module
- added D_() method for deferred translation
- languageSwitch method allow to dynamically change translation language
- import gettext is tested against ImportError, and dummy methods are used when not available (mainly useful for Libervia)
author | Goffi <goffi@goffi.org> |
---|---|
date | Sun, 29 Dec 2013 17:06:01 +0100 |
parents | a25db3fe3959 |
children | 1fe00f0c9a91 |
rev | line source |
---|---|
668 | 1 #!/usr/bin/python |
2 # -*- coding: utf-8 -*- | |
3 | |
4 # SAT plugin for Publish-Subscribe (xep-0071) | |
5 # Copyright (C) 2009, 2010, 2011, 2012, 2013 Jérôme Poisson (goffi@goffi.org) | |
6 | |
7 # This program is free software: you can redistribute it and/or modify | |
8 # it under the terms of the GNU Affero General Public License as published by | |
9 # the Free Software Foundation, either version 3 of the License, or | |
10 # (at your option) any later version. | |
11 | |
12 # This program is distributed in the hope that it will be useful, | |
13 # but WITHOUT ANY WARRANTY; without even the implied warranty of | |
14 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
15 # GNU Affero General Public License for more details. | |
16 | |
17 # You should have received a copy of the GNU Affero General Public License | |
18 # along with this program. If not, see <http://www.gnu.org/licenses/>. | |
19 | |
771 | 20 from sat.core.i18n import _ |
668 | 21 from logging import debug, info, error |
22 | |
23 from wokkel import disco, pubsub, iwokkel | |
24 from zope.interface import implements | |
25 # from lxml import etree | |
26 from lxml import html | |
27 try: | |
28 from twisted.words.protocols.xmlstream import XMPPHandler | |
29 except ImportError: | |
30 from wokkel.subprotocols import XMPPHandler | |
31 | |
32 NS_XHTML_IM = 'http://jabber.org/protocol/xhtml-im' | |
33 NS_XHTML = 'http://www.w3.org/1999/xhtml' | |
34 | |
35 PLUGIN_INFO = { | |
36 "name": "XHTML-IM Plugin", | |
37 "import_name": "XEP-0071", | |
38 "type": "XEP", | |
39 "protocols": ["XEP-0071"], | |
40 "dependencies": ["TEXT-SYNTAXES"], | |
41 "main": "XEP_0071", | |
42 "handler": "yes", | |
43 "description": _("""Implementation of XHTML-IM""") | |
44 } | |
45 | |
46 allowed = { | |
47 "a": set(["href", "style", "type"]), | |
48 "blockquote": set(["style"]), | |
49 "body": set(["style"]), | |
50 "br": set([]), | |
51 "cite": set(["style"]), | |
52 "em": set([]), | |
53 "img": set(["alt", "height", "src", "style", "width"]), | |
54 "li": set(["style"]), | |
55 "ol": set(["style"]), | |
56 "p": set(["style"]), | |
57 "span": set(["style"]), | |
58 "strong": set([]), | |
59 "ul": set(["style"]), | |
60 } | |
61 | |
62 styles_allowed = ["background-color", "color", "font-family", "font-size", "font-style", "font-weight", "margin-left", "margin-right", "text-align", "text-decoration"] | |
63 | |
64 blacklist = ['script'] # tag that we have to kill (we don't keep content) | |
65 | |
66 | |
67 class XEP_0071(object): | |
68 SYNTAX_XHTML_IM = "XHTML-IM" | |
69 | |
70 def __init__(self, host): | |
71 info(_("XHTML-IM plugin initialization")) | |
72 self.host = host | |
702
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
73 self.synt_plg = self.host.plugins["TEXT-SYNTAXES"] |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
74 self.synt_plg.addSyntax(self.SYNTAX_XHTML_IM, lambda xhtml: xhtml, self.XHTML2XHTML_IM, [self.synt_plg.OPT_HIDDEN]) |
668 | 75 host.trigger.add("MessageReceived", self.messageReceivedTrigger) |
702
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
76 host.trigger.add("sendMessage", self.sendMessageTrigger) |
668 | 77 |
78 def getHandler(self, profile): | |
79 return XEP_0071_handler(self) | |
80 | |
81 def _messagePostTreat(self, data, body_elt): | |
82 """ Callback which manage the post treatment of the message in case of XHTML-IM found | |
83 @param data: data send by MessageReceived trigger through post_treat deferred | |
84 @param xhtml_im: XHTML-IM body element found | |
85 @return: the data with the extra parameter updated | |
86 """ | |
87 #TODO: check if text only body is empty, then try to convert XHTML-IM to pure text and show a warning message | |
88 def converted(xhtml): | |
89 data['extra']['xhtml'] = xhtml | |
90 return data | |
702
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
91 d = self.synt_plg.convert(body_elt.toXml(), self.SYNTAX_XHTML_IM, safe=True) |
668 | 92 d.addCallback(converted) |
93 return d | |
94 | |
702
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
95 def _sendMessageAddRich(self, mess_data, profile): |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
96 """ Construct XHTML-IM node and add it XML element |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
97 @param mess_data: message data as sended by sendMessage callback |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
98 """ |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
99 def syntax_converted(xhtml_im): |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
100 message_elt = mess_data['xml'] |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
101 html_elt = message_elt.addElement('html', NS_XHTML_IM) |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
102 body_elt = html_elt.addElement('body', NS_XHTML) |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
103 body_elt.addRawXml(xhtml_im) |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
104 mess_data['extra']['xhtml'] = xhtml_im |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
105 return mess_data |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
106 |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
107 rich = mess_data['extra'].pop('rich') |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
108 syntax = self.synt_plg.getCurrentSyntax(profile) |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
109 d = self.synt_plg.convert(rich, syntax, self.SYNTAX_XHTML_IM) |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
110 d.addCallback(syntax_converted) |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
111 return d |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
112 |
668 | 113 def messageReceivedTrigger(self, message, post_treat, profile): |
114 """ Check presence of XHTML-IM in message | |
115 """ | |
116 try: | |
117 html_elt = message.elements(NS_XHTML_IM, 'html').next() | |
118 body_elt = html_elt.elements(NS_XHTML, 'body').next() | |
119 # OK, we have found rich text | |
120 post_treat.addCallback(self._messagePostTreat, body_elt) | |
121 except StopIteration: | |
122 # No XHTML-IM | |
123 pass | |
124 return True | |
125 | |
702
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
126 def sendMessageTrigger(self, mess_data, treatments, profile): |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
127 """ Check presence of rich text in extra |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
128 """ |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
129 try: |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
130 rich = mess_data['extra']['rich'] |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
131 # OK, we have found rich text |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
132 treatments.addCallback(self._sendMessageAddRich, profile) |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
133 except KeyError: |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
134 # No rich text found |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
135 pass |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
136 return True |
a25db3fe3959
plugin XEP-0071: rich messages management for sendMessage
Goffi <goffi@goffi.org>
parents:
701
diff
changeset
|
137 |
668 | 138 def _purgeStyle(self, styles_raw): |
139 """ Remove unauthorised styles according to the XEP-0071 | |
140 @param styles_raw: raw styles (value of the style attribute) | |
141 """ | |
142 purged = [] | |
143 | |
144 styles = [style.strip().split(':') for style in styles_raw.split(';')] | |
145 | |
146 for style_tuple in styles: | |
147 if len(style_tuple) != 2: | |
148 continue | |
149 name, value = style_tuple | |
150 name = name.strip() | |
151 if name not in styles_allowed: | |
152 continue | |
153 purged.append((name, value.strip())) | |
154 | |
155 return u'; '.join([u"%s: %s" % data for data in purged]) | |
156 | |
157 def XHTML2XHTML_IM(self, xhtml): | |
158 """ Convert XHTML document to XHTML_IM subset | |
159 @param xhtml: raw xhtml to convert | |
160 """ | |
161 # TODO: more clever tag replacement (replace forbidden tags with equivalents when possible) | |
162 | |
163 parser = html.HTMLParser(remove_comments=True, encoding='utf-8') | |
164 root = html.fromstring(xhtml, parser=parser) | |
165 body_elt = root.find('body') | |
166 if body_elt is None: | |
167 # we use the whole XML as body if no body element is found | |
168 body_elt = html.Element('body') | |
169 body_elt.append(root) | |
170 else: | |
171 body_elt.attrib.clear() | |
172 | |
173 allowed_tags = allowed.keys() | |
174 to_strip = [] | |
175 for elem in body_elt.iter(): | |
176 if elem.tag not in allowed_tags: | |
177 to_strip.append(elem) | |
178 else: | |
179 # we remove unallowed attributes | |
180 attrib = elem.attrib | |
181 att_to_remove = set(attrib).difference(allowed[elem.tag]) | |
182 for att in att_to_remove: | |
183 del(attrib[att]) | |
184 if "style" in attrib: | |
185 attrib["style"] = self._purgeStyle(attrib["style"]) | |
186 | |
187 for elem in to_strip: | |
188 if elem.tag in blacklist: | |
189 #we need to remove the element and all descendants | |
190 debug(u"removing black listed tag: %s" % (elem.tag)) | |
191 elem.drop_tree() | |
192 else: | |
193 elem.drop_tag() | |
701
98b2400e17d6
plugin XEP-0071: XHTML2XHTML_IM don't return the <body> root tag anymore.
Goffi <goffi@goffi.org>
parents:
668
diff
changeset
|
194 if len(body_elt) !=1: |
98b2400e17d6
plugin XEP-0071: XHTML2XHTML_IM don't return the <body> root tag anymore.
Goffi <goffi@goffi.org>
parents:
668
diff
changeset
|
195 root_elt = body_elt |
98b2400e17d6
plugin XEP-0071: XHTML2XHTML_IM don't return the <body> root tag anymore.
Goffi <goffi@goffi.org>
parents:
668
diff
changeset
|
196 body_elt.tag = "p" |
98b2400e17d6
plugin XEP-0071: XHTML2XHTML_IM don't return the <body> root tag anymore.
Goffi <goffi@goffi.org>
parents:
668
diff
changeset
|
197 else: |
98b2400e17d6
plugin XEP-0071: XHTML2XHTML_IM don't return the <body> root tag anymore.
Goffi <goffi@goffi.org>
parents:
668
diff
changeset
|
198 root_elt = body_elt[0] |
668 | 199 |
701
98b2400e17d6
plugin XEP-0071: XHTML2XHTML_IM don't return the <body> root tag anymore.
Goffi <goffi@goffi.org>
parents:
668
diff
changeset
|
200 return html.tostring(root_elt, encoding='unicode', method='xml') |
668 | 201 |
202 class XEP_0071_handler(XMPPHandler): | |
203 implements(iwokkel.IDisco) | |
204 | |
205 def __init__(self, plugin_parent): | |
206 self.plugin_parent = plugin_parent | |
207 self.host = plugin_parent.host | |
208 | |
209 def getDiscoInfo(self, requestor, target, nodeIdentifier=''): | |
210 return [disco.DiscoFeature(NS_XHTML_IM)] | |
211 | |
212 def getDiscoItems(self, requestor, target, nodeIdentifier=''): | |
213 return [] |